26–28 sept. 2022
APC, Paris
Fuseau horaire Europe/Paris

Data/MC adaptation with adversarial training in KM3NeT

28 sept. 2022, 09:40
20m
Amphithéatre Pierre Gilles de Gennes (sous-sol) (APC, Paris)

Amphithéatre Pierre Gilles de Gennes (sous-sol)

APC, Paris

4 rue Elsa Morante, 75013 Paris
2 ML for analysis : event classification, statistical analysis and inference, including anomaly detectio Wed morning

Orateur

Joao Coelho (APC / CNRS)

Description

Data and MC represent different domains. Supervised learning in MC needs to be transferred to the real data domain and MC mismodelling can reduce the performance of the transferred models. In this work we are implementing the Domain Adversarial Neural Network (DANN) concept to the standard Graph Neural Network (GNN) classification and regression tasks in the KM3NeT/ORCA experiment. We will present the current status and prospects for the project.

Auteurs principaux

Joao Coelho (APC / CNRS) Santiago Pena Martinez (Aix-Marseille Université) Shen Liang

Documents de présentation